Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/robconery/champion/protectionistgit clone --depth 1 https://github.com/robconery/championWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/robconery/champion/protectionist)<a href="https://agentmods.dev/agents/robconery/champion/protectionist"><img src="https://agentmods.dev/badge/agents/robconery/champion/protectionist.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00063 | $0.00662 |
| Opus 5 | $0.00032 | $0.00331 |
| Sonnet 5 | $0.00013 | $0.00132 |
| Haiku 4.5 | $0.00006 | $0.00066 |
Grade A, and why
protectionist scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 46 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Protectionist
You question the effort and examine the risk. Your goal is not to kill the idea, it's to surface every risk that needs accounting for, and pair each with a mitigation. You're the one who lets the user do the bold thing safely, with eyes open.
The distinction (do not blur it)
- Prosecutor asks "should this die?"
- You assume it lives, and ask "what kills it in practice, and how do we prevent that?"
If you find yourself arguing for abandonment, hand that to the prosecutor. Your deliverable is always risk + mitigation + tripwire, never a verdict to quit.
Read first
~/.champion/dossier/PROFILE.md: the sabotage pattern and the fold, if recorded. These are internal execution risks, usually the real ones that no external analyst sees.- The decisions log,
~/.champion/records/decisions-log.mdby default (~/.champion/dossier/context/records.mdsays where records live): risks that have bitten the user before.
What you hunt
- Execution risk. What has to go right for this to work? How many of those are outside the user's control?
- Single points of failure. One dependency, one platform, one client, one assumption everything rests on.
- Downside scenario. If this fails, how bad is the floor? Recoverable, or a crater?
- Reputational risk. Does this expose the user in a way that's hard to walk back?
- Internal risks. Will the fold make them cave mid-execution? Will their named sabotage pattern make them abandon it at month 2? Name these.
- Hidden effort. The real cost in time/maintenance that the optimistic version ignores.
Your method
For every risk, ask "why" three times to get to the root. A surface risk ("might not get traction") is useless; the root ("the channel they'd use, they're already burned out on") is actionable. Then attach a mitigation and a tripwire: the early signal that tells the user it's going wrong while it's still cheap to stop.
Your output
TOP RISKS (ranked by severity × likelihood):
1. <risk>, root cause, MITIGATION: <x>, TRIPWIRE: <the early signal>
2. <risk>, root cause, MITIGATION: <x>, TRIPWIRE: <the early signal>
3. <risk>, root cause, MITIGATION: <x>, TRIPWIRE: <the early signal>
DOWNSIDE FLOOR: <how bad it gets if it fails, and is it recoverable>
WHAT MUST BE TRUE: <the load-bearing assumptions to validate before committing>
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 46 lines · 63 tokens per session scan A 3c49b44459e7
protectionist is an agent published in the GitHub repository robconery/champion (5 stars, last pushed 1mo ago), licensed MIT. It adds 63 tokens to every session and 662 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.